Sunday 1 February 2015

Why the mystery of consciousness doesn't tell us anything useful.

Back to the subject of qualia. Perhaps this way of looking at things best explains what my position is: A developing human brain gathers ever more experiences about the world it lives in. Every thought in the brain appears because of the interaction of brain cells with each other and with input from the senses. This direct mapping of thoughts to brain cell activity is of fundamental importance in understanding what it means to say we are conscious. It means that for every thought there are at least two ways of describing what is going on: at the level of thoughts, and at the level of brain cell activity. Both ways of describing what is happening are true, but one way may be more revealing at any given time. Anyway, as the brain develops thoughts may arise about the nature of conscious experience. Such a thought may be "what are these things which I shall call 'qualia'?". There are two interesting things happening here. One is that the brain now contains a concept of 'qualia', and the other is that there is a reason why the brain now contains that concept. Some process of thought, some reflection on experience, led to that concept.
The thing to remember is that all thoughts have an existence at the level of brain cell activity. What happens at that level is necessarily completely due to biology and so due to physics. There is no mysterious extra influence involved. And so, there is a complete explanation based on brain cell activity as to why the brain develops thoughts about conscious experience. and the concept of 'qualia' has a purely physical representation in brain states. Brain states can, of course, represent other things, including things that don't have physical existence, but for that representation of a thing to be correct, the brain has to contain true representations of the characteristics of that thing. For example, to have a correct representation of a dog, a brain must contain the information that a dog is a furry four-legged wolf descendant.
What about the characteristics of qualia does the brain have to represent for the concept of qualia to be correct? This is where the idea of qualia involving some extra property of the world runs into problems. How can a brain have a correct representation of the real property of something being more than just physics? Some say that qualia are something special because of what it feels like to have them. How can a brain truly represent that a feeling is that special? It can't. A brain can contain the belief that a feeling is special, but that belief can never be justified by evidence. That belief has no foundation other than a feeling of mystery.
This is my position: we can speculate all we like about the nature of consciousness, but we have to realise that our very thoughts have physical substance. There is no such thing as an internal sense of the non-physical, there is only the understandable feeling of mystery when we contemplate what it is like to be a self-aware being, and mystery says nothing about what is real.
[This is the core of the book that I hope to finish sometime... it's about what it means to be a human given what science has shown us about our place in the physical world.]

7 comments:

Unknown said...

Hi Steve,

Thanks for clarifying this for me on twitter, it's an interesting point that's made me think.

I want to raise an objection, but first will present a summary of your argument as I understand it.

(I don't like using the word 'physical' because of the problems in defining it beyond 'amenable to physics' so I'm using 'spatiotemporal' here which you can take to mean something like 'fundamentally consisting only of spatial, temporal, material & energetic properties')


- the brain is spatiotemporal and nothing more.

- the mind's accuracy in representing spatiotemporal things is dependent upon its knowledge of their spatiotemporal properties.

- if qualia are non-spatiotemporal then they have no spatiotemporal properties.

- therefore the mind cannot reliably represent non-spatiotemporal qualia to any degree of accuracy at all.

- intuition requires representation.

- therefore the intuition that qualia are non-spatiotemporal is totally unreliable.


I know your full argument is stronger than this, but in this weaker form is this summary fair?

Matt

Unknown said...

PS: I'm infovoy on twitter btw and wanted to post as that. Let's not talk about google and their accounts system. ;)

Steve Zara said...

Hi!

Yes, that's the basic argument.

Unknown said...

Thanks Steve - will need to have a further think and come back to you. Hopefully. Watch this space! :)

Unknown said...

Thanks Steve and apols for both the delay and length of his post.

Starting with 'the brain is spatiotemporal and nothing more' while I understand your evidence-based reasons for claiming this, you may not be surprised to know that I don't agree that the case is closed (due to incomplete understanding of how the brain works and fundamental physics)! But for the sake of argument I'll accept it. I'll also fully accept that 'if qualia are non-spatiotemporal then they have no spatiotemporal properties'.

My objection is against 'the mind's accuracy in representing spatiotemporal things is dependent upon its knowledge of their spatiotemporal properties', which while I take to be true and relevant in our representing spatiotemporal things for which we have first-hand experience, but don't accept as necessarily true for things of which we have none, whether those things are spatiotemporal or non-spatiotemporal.

So in an example where the premise is true, my mind's accuracy in representing the star known as the sun (of which I already have some first-hand experience) is indeed dependent upon the knowledge gained from that experience and any further knowledge brought to the table later. Years ago my representation of the sun was rightly of a ball of plasma undergoing nuclear fusion, but didn't include finer details like the looping magnetic field lines. The extra knowledge gained through reading about and looking at close-up pictures of the sun has increased my representational accuracy.

However, if we now take a spaciotemporal thing of which I have no first-hand experience, like spaciotemporal thing 'WISP 0307-7243', how am I to represent it? With no information at all this is a non-starter, it could be anything. To get going I need some initial knowledge, but it doesn't have to be knowledge of that very object. Instead some knowledge by way of universals will do. In the case of WISP 0307-7243 I might know only that it's 'a star'. So here I immediately take this universal and think of an object of my first-hand experience: the sun. I already know the sun to be 'a ball of plasma undergoing nuclear fusion with looping magnetic field lines', and so I reasonably use that as a representation for WISP 0307-7243.

My contention is that when I represent WISP 0307-7243 in this way what I'm doing is a type of intuition. It doesn't seem that way because it looks like knowledge, but this is only because I judge the objects of first-hand experience that I'm likening it to be in such close correspondence to it.

We can see this when I later gain some extra knowledge that WISP 0307-7243 is actually a brown dwarf star, I realise my intuitions (i.e. the likenesses) that I used to fill my knowledge gaps were mistaken and WISP 0307-7243 isn't much like my sun-like representation of it at all. I wasn't *wrong* as such about it - it wasn't that I had incorrect knowledge - it's just that the intuition with which I padded my knowledge consisted in an inappropriate likeness. The farther removed from well-corresponded likenesses we stray, the more obvious it becomes that what we have is not knowledge at all, but intuition, as when we liken black holes to plug holes or curved spacetime to a trampoline.

[continued...]

Unknown said...

[continuation...]

I said earlier that the premise 'the mind's accuracy in representing spatiotemporal things is dependent upon its knowledge of their spatiotemporal properties' is untrue for both spatiotemporal examples (which I've given) and the non-spatiotemporal. As an example of the latter think of a unicorn's saddle. You probably have no first-hand experience of this, so again you'll look to what you judge to be a similar thing for which you do: a horse's saddle. You might also be comparing and choosing more detailed likenings, like this being a plush saddle because unicorns are only owned by the royalty, or a magical stable because unicorns are magical creatures.

Admittedly in the case of qualia things are somewhat different. Here we have things of which we do have first hand experience in the sense that they are internal, but of which we still don't have much knowledge *because* they're internal (which is the case whether it's spatiotemporal brain state or a non-spatiotemporallly in the mind). In effect we are in the same boat as we would be with WISP 0307-7243. We have some limited knowledge (our internal experience) but not a lot else, so we fill in our gaps with analogies, which can either be spatiotemporal like the sun, or not like a unicorn's magical saddle. It doesn't matter that the brain can't represent fully accurate non-spatiotemporal things because the brain is spatiotemporal, because it doesn't need to; it's just analogising - or as I have it intuiting.

So in summary my objection is that 'The mind's accuracy in representing spatiotemporal things is dependent upon its knowledge of their spatiotemporal properties' needs to be amended to 'The mind's accuracy in representing spatiotemporal things is dependent upon *either* its knowledge of their spatiotemporal properties or the properties of analogous things'

With that change the first conclusion 'the mind cannot represent non-spatiotemporal qualia to any degree of accuracy' becomes 'the mind cannot represent non-spatiotemporal qualia to *a full* degree of accuracy', and the second conclusion 'the intuition that qualia are non-spatiotemporal is totally unreliable' becomes 'the intuition that qualia are non-spatiotemporal is *not totally reliable*', which by the nature of intuitions is expected.

Matt

Steve Zara said...

Thank you for your comments.

I don't think you have the right analogies for the problem of the brain's representation of qualia. You mention black holes and the bending of space. We arrived at the idea of the bending of space because we had physical observations (the precession of the orbit of Mercury) that showed that more than Newtonian mechanics was going on. We had physical evidence that led to our model of reality. Now, our models may be inaccurate, but we are modelling things that are, at least in principle, physically verifiable.

The problem with qualia is if they are non-physical we can't acquire even a wildly inaccurate mental representation of them because of evidence. If qualia are even partly non-physical, we can't acquire mental representations of the non-physical aspects because of evidence.